Millions of Lafferty Words and Intellectual Compression

"Throughout his writing career, these various affects and effects were used by Lafferty to construct stories and novels that nestled within larger (but often untold) tales and universes, and were often to be understood as epigonal offshoots of those larger, earlier, linguistically more complex, mysteriously governed worlds. In the end, his corpus as a whole gave off a sense of teasing incompletion and of secrecy: almost as though it was only the entire story (to which his numerous unpublished manuscripts added almost mythic stature) that made sense, that commanded the rest." SFE

Some thoughts from the weekend, as I found myself reconsidering that stylish, romantic, yet demystifying passage by John Clute. The Oxford don and book reviewer John Carey once wrote that any fine book review must contain at least one such passage. He called it a poison pill. Without its bewitchment, one has not read a good review. This is what the 20th-century rhetorician Kenneth Burke called rhetorical form: the satisfaction of a desire. Clute suggests what one experiences in Lafferty is a trompe-l'œil. What do I mean? And how is that a poison pill?

After reading the SFE entry passage, you might come away feeling smarter than you actually are. After all, you might recognize a readerly experience you have had in reading something like the The Devil is Dead, an alienated thought expressed better than perhaps you could have expressed it. Perhaps you sense that you have been one of these teased readers. You have mistaken your reader's hope that a larger picture exists for the thing itself. There may, in fact, be no such thing. There is (and here is the poison in the pill) only teasing incompletion.

I like rhetoric. In rhetorical terms, this is masterfully enthymematic: you are led to draw your own conclusion, having been mobilized to draw out the implied but unsaid.

P1: If individual works appear to point toward larger, hidden structures, then those structures must exist and must contain the coherence of the author's vision.

P3 (stated/implied): Lafferty’s stories and novels appear to be shards, offshoots, or epigones of larger, untold, more complex worlds.

P3 (supplied by the reader): Therefore, there is a single, vast, unified “entire story” behind the works, a totality that commands and explains them.

The enthymeme lays out the logic the reader is tempted to follow. Clute activates that P3 inference and, with his final emphasis on teasing incompletion, undercuts it. That sense of an overarching structure? It might be a mirage, an effect of Lafferty’s writing, not a good reason to suspect a coherent architecture.

In Carey’s poison-pill terms, if you have activated the enthymeme and went with it being undercut, you swallowed the poison pill. You might feel satisfied. You have just read the equivalent of a fine book review, which satisfies the need to read the actual book, a book review being a faster, cheaper substitute form for the attentional investment of reading a book. There isn’t enough time to read all the books, after all, which is why book reviews exist. Many poison pills are medicinal, even when they are lies.

There is nothing wrong with this, and I appreciate the master poisoners in the book trade. Most books deserve to be poisoned like vermin. In this case, Clute suggests that the Lafferty key does indeed turn, but that it may open onto nothing larger than what is already visible. This is the unpublished work’s “mythic stature.” Mythic here is not the good mythic; it is, in part, the other kind of mythic, like Queen Elizabeth’s mythic virginity. Perhaps it would simply be more Lafferty. Or perhaps Lafferty is best taken as a handrail that only ever seemed to be leading somewhere. Or perhaps Clute does not say any of this directly, because one never states a missing premise of an enthymeme—the rhetorician makes you supply it. Maybe Clute is merely articulating what you experienced as a reader. And what did you experience? That is how an enthymeme works. Trust yourself after trusting me to show you the best seat in the house.

Lafferty wrote millions of words, and he did so with the mind of an engineer. By his own account—and there is little reason to doubt him—he possessed something close to eidetic recall for written language. He compressed information. His mental audio was nearly lossless, and anyone who reads him will be struck, again and again, by how unerringly he could summon the exact word for whatever he described. Much of his pragmatic marker/oral style is meant to hide this finicky precisionism.

Among genre writers, the number of words he used only once is astonishing (see my post on the statistic for the publishing’s works)—never pretentious, always precise—and a clear sign that something in his stylistic machinery was fundamentally different. And then there is the matter of how thoroughly he hides writerly labor. This, strangely, is almost never remarked upon. Fun fact. Lafferty used more than nine hundred sources while writing The Fall of Rome, yet he took great care to conceal that scaffolding from the reader’s experience. The most any one has said or apparently noticed is Darrell Schweitzer, who talks about one or two.

One of my aims for this blog is to confront the two types of problems that arise when trying to read Lafferty at scale instead of in Best of X terms. For most Lafferty readers, he is a short story writer, someone who turns out brilliant, self-contained performances. Part of his genius as a writer is that he was able to present himself this way, and he is thoroughly enjoyable if one wants to read him as science fiction’s master vaudevillian. At the same time, as I’m sure is clear if you read this blog, it is a bit of a put-on. The man was up to other things. When someone says he did not take himself seriously, it is true. When someone says that implies that he did not take his ideas seriously, it is bullshit, as reading his correspondence quickly reveals.

The problem for anyone wanting to get a grip on Lafferty is that he has two radically different strategies, and both create different sorts of reading pleasure. The is a modal gap between them. Both turn on the problem of informational density in his work. There is the centripetal Lafferty and the centrifugal Lafferty. Over the last few months, I’ve decided that the best way to talk about a Lafferty short story is just to call it a story world. I thought about calling these story worlds cosmiums or something similar because they tend to have their wild logics and metaphysical rules, and part of Lafferty’s gambit when he writes a story is to make each cosmium (aside from the story sequences) both continuous and discontinuous with the others. On the continuous side, you have the ghost story, with its lines of filiation, overlapping themes, and reduplicative conceptual patterning. I needed new concepts to see this; hence, the concepts page. On the other hand, there is the discontinuous, where each story world has its own conceptual architecture, dense networks of allusion, etymological texture, anti-secrets, and unresolved puzzles, along with more than a little didactic axe-grinding. What ties the two sides together, the continuous and discontinuous, is informational density. When most himself in his short fiction, Lafferty is an easy-going but camouflaged maximalist.

The writers I tend to know best and enjoy most are maximalists. I believe the universal appears in the particular, so the more detail an a literary artist can control, the better. It is one reason why the writers I’m probably most knowledgeable about are Dante, Milton, Blake, and Joyce. Lafferty is becoming one of those writers for me, but this is something of an accident. Other writers I love could command this kind of effort. I’m a big fan of William Gaddis and William T. Vollman, and I can see doing something similar with their canons—trying to set out some maps. One of my heroes is Donald Ault who knew the important street contained both both Duckburg and The Four Zoas. It is not trying to raise low art into high art. I think the same would benefit one of my favorite terrible writers, the mesmerizing Harry Stephen Keeler (1890-1967). (As an aside, one of the great books for reading someone like Keeler or Lafferty is the forgotten early work on the artistry of schizophenics written by Hans Prinzhorn (1886-1933), Artistry of the Mentally Ill, not that Keeler or Lafferty were mentally ill; rather, both used phantasmatic artistic strategies for depicting pathological culture). What writers like this have in common is that they are architectonic. Lafferty has the oceanic; Keeler has his webwork. Lafferty is peculiar in that he is “outsider” architectonic, as is Keeler, but his readers have not read him that way. My view is that it comes down to the contingencies of publication history, his developmental pattern, and his intense privacy, which was combined with being deeply out of sync with his time, as most visionaries always are.

There are now new tools in the banner intended to help a reader (me) better grasp Lafferty’s informational density, which brings me to the point of this post. I think Lafferty is a writer who is uniquely interesting in terms of where the digital humanities are now. In the past, one could not have expended the amount of energy necessary to get the critical apparatus in place to read Lafferty thoroughly because he is, in terms of reputation if not in terms of merit, critically negligible. In the past, I have written a little bit about George Steiner’s ideas of difficulty and why they seem relevant to me in reading Lafferty. The four kinds of difficulty are always on my mind when reading hard material, so here they are again. What follows is how I see the issue now, along with the challenges and promises that Lafferty poses for dogged readers.

Contingent Difficulty: Reconstructing Lafferty’s Informational DensitySteiner’s contingent difficulty is really an idea that identifies works that require external knowledge the reader does not yet possess, and Lafferty scores high here. First, he is profoundly literate theologically in a tradition that most of his readers will not understand, then his fiction is saturated with etymological play, what I have called historical bricospolia, metaphysical argument, lost theological debates, Indigenous histories, displaced quotations, and private symbolic systems that are going to exceed any single reader’s grasp. What once required decades of archival work can now be approached with AI-powered concordances, semantic search, and automated pattern detection. I use these tools in my thinking about Lafferty because I think they are here to stay and because Lafferty is an astonishing test case for them. The rise of these tools and the accident of having become interested in Lafferty at the same time has been an important learning experience for me. These tools do not replace interpretation, which is what my blog posts do, but they do make it possible to assemble the distributed information needed to even begin reading Lafferty macroscopically. Digital humanities methods (vectorized text similarity, clustering, allusion detection) allow the reader to model the informational density that previously overwhelmed traditional close reading. This doesn’t demystify Lafferty. In a way, it restores the conditions of uptake, making apparent the contingent frameworks that his fiction assumes. Conceptual shortcuts, as it were.

Modal Difficulty: Adjusting Our Reading Practices to Postfictional Forms. When I teach Dante to students, one of the biggest challenges is helping them overcome the hurdle of modal difficulty, especially in the final part of The Divine Comedy, Paradiso. Steiner’s modal difficulty arises when an artistic work demands that readers adjust their expectations of what literature is. Lafferty’s really big claim that the novel is a historically bounded form—and that he writes prose fiction after the last novels—means his long works must be read through a different interpretive mode. Frankly, I think this is why people don’t get his novels beyond the books of 1968 and 1969. They resemble novels in length but are not novels in kind; they participate in Romance, theological comedy, metaphysical case study, eschatological reportage, and mock-scholastic fabulation. They are doing all kinds of things that look more like what happened before the novel took on its form in the 18th century than after it. In a sense, Lafferty’s prenucleation to ghost story leap is a leap past the novel form that happened because the novel form was a failure for him, both the serious novel, as with Civil Blood, and the more mainstream genre novel, as in Mantis. AI-assisted modeling can help clarify these modal shifts by mapping narrative structures, quick scene changes, and intertextual genealogies across his corpus. These tools can help someone see what entrenched reading expectations miss: that Lafferty is building architectures of meaning whose logic is closer to Dante, patristics, Indigenous cosmology, medieval summae, and the satires of early modernity than to 19th and 20th-century fiction. Most science fiction writers were parasitic on those forms, but Lafferty became increasingly cranky about that. He called it heroic tedium. The digital humanities allow us to visualize these modal transformations.

Tactical Difficulty: Lafferty’s Deliberate Obliqueness and AI as a Pattern-Tracking Supplement. Steiner’s third form of difficulty is tactical difficulty—difficulty by design—and it is everywhere in Lafferty. He doesn’t expect the reader to work, but he expects his ideal reader to be smart and to have worked. He obscures transitions, inverts causality, screws with conventions, and embeds metaphysical premises without explication. He rejects the conventional notion of the novelistic character by going back to the older concept of character as type, and then he treats the pre-modern type as being fractal, as if type were explosive and radiating outward. This difficulty is, again, architectonic, forming the latticework of the Ghost Story. Traditional interpretation can easily describe local narrative tactics, but AI tools excel at detecting recurrent structural moves: the recurrence of PRIME-coded metaphysical cues, the circulation of symbolic numbers, the self-replicating narrative geometries, the fractal reappearance of themes across short stories and long works. One of the things that I don’t like about the way most people have blogged about Lafferty is that they have tended to say you can understand Lafferty by understanding this more easily understandable, more conventionalized, less difficult form. The fact that Lafferty’s tall tales are cosmiums is probably the biggest exemplar of this problem. When someone says, here is a tall tale, it’s usually a sign that the person is skirting the informational difficulty of the story world.

Ontological Difficulty: PRIME, Metaphysical Commitment, and the Limits of Computation. This is deep Lafferty, and it is Steiner’s deepest form of difficulty—ontological difficulty. It happens when a text challenges the conditions of meaning themselves. Frankly, this is the center of Lafferty’s project. Someone like Petersen is right to recognize that the stakes are ontology, not science fiction. Lafferty’s clear-eyed view on what he took to be PRIME, the real world (whatever he got wrong about it), does determine the metaphysical substrate against which all post-novelistic fiction must answer. It means that his writing scribbles at the boundary where narrative becomes metaphysical inquiry. Digital humanities tools can trace the distribution of metaphysical motifs and model cross-textual topographies, but they cannot resolve ontological commitments or adjudicate metaphysical truth. That is where people need to argue. However, they can help us perceive how Lafferty sets up this difficulty: how historical mimēsis, Romance, metaphysical comedy, and fabulation all refract the problem of how contingent worlds relate to the World. And because the final difficulties are always ones of ontological commitment, the new tools of the digital humanities stop here, right where the ontology of the person begins.